Académique Documents
Professionnel Documents
Culture Documents
ABSTRACT
The long-term creep strength of the new generation of martensitic creep resistant 9-12%Cr
steels since the well-known steel Grade 91 relies strongly on particle strengthening by fine
MN nitrides based on V and Nb. However, during long-term high-temperature exposures the
MN nitrides may be replaced by the thermodynamically more stable Z-phases (Cr(V,Nb)N).
Since Z-phase precipitates as coarse particles, which do not contribute to the creep strength in
the same way as the fine MN nitrides, Z-phase precipitation causes a breakdown in creep
strength. Study of the Z-phase precipitation process including electron microscopy and
development of a thermodynamic model has led to increased understanding. A strong
correlation between Cr content and Z-phase precipitation rate is now evident: In steels with
around 9%Cr Z-phase precipitation is too slow to disturb the MN population within 100,000
hours at 650 °C, but in steels with Cr contents above 10.5% Z-phase precipitation is strongly
accelerated. This effect explains the lack of success for all attempts to develop MN
strengthened martensitic creep resistant steels with Cr content around 12% for oxidation
protection up to 650 °C. The developed thermodynamic model predicts the influence of
several alloying elements, which enables control of the Z-phase precipitation rate. This has
now been used for innovative alloy design: By strong acceleration of the precipitation process
a fine dispersion of small stable Z-phases can be generated, which may contribute to creep
strength in the same way as the MN nitrides usually do. This work has led to alloy design of
experimental 12%Cr martensitic steels strengthened by Z-phase. Such steels may again enable
the combination of high strength and oxidation resistance in the same alloy, and open a new
pathway for further alloy development of the heat resistant martensitic steels.
1. Introduction
Because of their combination of high creep and fatigue strength and moderate cost, the
martensitic 9-12%Cr steels are vital materials for further increases in steam parameters of
fossil fired steam power plants. Recently, new stronger steels like the 9%CrW steel Grade 92
for thick section boiler components and steam lines, and the 9%CrCoB steels FB2 and CB2
for large forgings and castings, have been introduced in commercial power plant projects.
These steels have enabled the construction of power plants with supercritical live steam
parameters up to 600°C/300 bar [1,2]. The improved efficiency of these new plants compared
to earlier subcritical steam plants (540°C/180bar) corresponds to a reduction of ca. 30% in
specific CO2 emission [3].
In order to further increase the steam parameters of steel based power plants up to a target
value of 650°C/325 bar it is necessary to double the creep strength compared with Grade 92,
and at the same time the resistance against steam oxidation must be improved. If the oxidation
protection is to be achieved through alloy additions instead of surface coatings, it is necessary
to increase the Chromium content in the steels from 9% to 12% [4]. However, so far all such
attempts to make stronger 12%Cr steels have been unsuccessful because the high chromium
content introduced severe microstructure instabilities in the tested steels, which led to
breakdowns in long-term creep strength [5].
The long-term creep strength of the new generation of martensitic creep resistant steels since
the well-known steel Grade 91 relies strongly on particle strengthening by fine MN nitrides
based on V and Nb. Additions of ca. 0.15-0.25%V, 0.04-0.08%Nb and 0.02-0.07%N are
common to all of the new steels, due to results obtained by Fujita in the late 1970s [6].
However, recent research has demonstrated that the MN nitrides, (V,Nb)N, may be replaced
by the thermodynamically more stable Z-phases, Cr(V,Nb)N, which precipitate as coarse
particles and dissolve the fine nitrides [7,8]. This phase transformation has been found to be a
main cause for observed long-term microstructure instabilities in the new martensitic creep
resistant steels. Thus, the Z-phase transformation is a main cause for the lack of success to
develop strong martensitic steels with 12%Cr for improved oxidation resistance.
In the present paper the Z-phase transformation is described in detail, and its influence on the
long-term stability of 9-12%Cr steels is discussed. Finally, a possible way to overcome
present difficulties to combine high Cr content with long-term microstructure stability is
outlined.
The 9-12%Cr steels have mainly tempered martensite microstructures formed during a final
normalising and tempering heat treatment, although for some compositions considerable
amounts of δ-ferrite may form in the microstructure. The tempered martensite consists of
prior austenite grains separated by high angle grain boundaries. These grains are subdivided
by high angle boundaries into blocks of martensite laths [9]. After tempering the blocks
consist of subgrains with an average width of 0.3-0.5 μm separated by low angle boundaries.
The dislocation density of the tempered martensite is high (in the order of 1014 m-2), Fig. 1.
The technically interesting stress and temperature ranges for creep testing and service
exposure of the 9-12%Cr steels are 250MPa-20MPa at 500°C-700°C. In these ranges the main
creep mechanism in the steels is dislocation creep. This leads to glide of free dislocations and
migration of subgrain boundaries during creep, resulting in a reduction of the dislocation
density and growth of subgrains [10,11]. These deformation processes can be delayed by
precipitate particles, which pin dislocations and subgrain boundaries.
A number of different precipitates can be found in the 9-12% Cr steels. Depending on their
particle hardening effect and stability against coarsening or dissolution during long-term
exposure the precipitates determine the microstructure stability of the steels. Precipitate
hardening should be regarded as the most significant creep strengthening mechanism to obtain
high long-term creep strength in the 9-12%Cr steels [12,13]. The most useful precipitates are
the Cr carbides M23C6, the intermetallic Laves phases Fe2(Mo,W) and the MN nitrides
(V,Nb)N. The MN nitrides precipitate in high number densities as very fine particles, which
are much more stable against coarsening than the other precipitates [14]. As such they
contribute significantly to long-term precipitation hardening in the steels. Boron addition to
the steels may stabilise the M23C6 carbides, especially near the prior austenite grain
boundaries. [15,16].
Z-phase Cr(V,Nb)N was first found in a martensitic 11%Cr steel by Schnabel et.al. in 1987
[17]. In 1996 Vodarek and Strang realised that Z-phase precipitation could explain observed
sigmoidal creep behaviour in older 12CrMoVNb steels, since the Z-phase precipitates as
coarse particles and dissolves fine MN nitride particles [8].
Recently, the authors made systematic studies of Z-phase content in a number of 9-12%Cr
steels, which rely on MN strengthening [7]. Both commercial and experimental grades were
investigated after exposure to creep, and a clear correlation between the Cr content of the
steels and the observed Z-phase quantity was found, table 1. Comparisons with creep testing
on a number the investigated steels show that for Cr contents above 10.5% the strongly
accelerated Z-phase precipitation leads to a breakdown in long-term creep strength steels. In
steels with Cr contents below 9% Z-phase precipitation is so slow that they are largely
unaffected up to very long testing times at 600-650°C [7].
Results from detailed studies of the mentioned steels were used to develop a thermodynamic
equilibrium model of the Z-phase [18]. This model has demonstrated that Z-phase Cr(V,Nb)N
is thermodynamically the most stable nitride in all 9-12%Cr steels containing V, Nb and N,
see Fig. 2. This means that Z-phase should eventually form in all of the new generation 9-
12%Cr steels with potentially detrimental effects on their creep strength. However, as seen
from table 1, large differences in Z-phase precipitation rate has been found, which seem to
correlate well with similarly observed large differences in the rate of breakdown in creep
strength.
The thermodynamic equilibrium model can be used to estimate Z-phase precipitation rates by
calculations of the thermodynamic driving force for the Z-phase precipitation. This driving
force is the difference in free energy between the steel with all MN nitrides present and the
steel with only Z-phase present, and it can give indications of the influence of chemical
composition on the Z-phase precipitation rate. Such calculations clearly demonstrate that Cr is
the most influential element for the Z-phase driving force, see Fig. 2. Further it is found that
elements like C or Co, which are not included in the Z-phase, also affect its driving force. C
forms as M23C6, which ties up Cr and thus effectively lowers its content, while Co reduces the
Cr affinity in the matrix, thus encouraging it to form as Z-phase. Results by Strang and
Vodarek indicated a strong accelerating effect of Ni content on Z-phase precipitation [8].
However, the model studies provide no support to any effect of Ni content on the precipitation
rate of the Z-phase.
Figure 4. Large (a) and small (b) Z-phase particles co-existing in a 10.5%Cr steel after 43,000
hours at 600°C.
3. The Z-phase precipitation process
Even though the Cr(V,Nb)N Z-phase is thermodynamically the most stable nitride in 9-
12%Cr steels alloyed with V, Nb and N it seems quite difficult for it to nucleate. Z-phase has
never been observed in any of the steels directly after tempering for a few hours in the normal
range 650-800°C. Even in the high Cr steels, where Z-phase forms most rapidly, widespread
Z-phase formation has only been observed after quite long exposure times of 1000 hours or
more at 650°C. Experimental studies of Z-phase TTP diagrams indicate that the fastest
precipitation happens at 650C [23].
Once a Z-phase has nucleated it can grow fast, and large Z-phase particles can be observed
after relatively short exposures. The Cr(V,Nb)N Z-phase grows by dissolution of the MN
precipitates, which provide V, Nb and N, and by picking up Cr from the steel matrix. In spite
of the fast growth, observations of long-term exposed specimens always reveal both large and
small Z-phase particles, Fig. 4. This indicates that the nucleation of Z-phase is slow and
continuous, and thus the nucleation process should be regarded as rate controlling for the Z-
phase transformation [24,25]. The fact that Z-phase grows fast means that even in early stages
of the transformation large Z-phase particles can be observed in the microstructure even
though only a very limited part of the MN population has been dissolved. Thus, an
observation of a few large Z-phase particles in the steel microstructure is not necessarily
alarming for the creep stability. Instead, attention should be focused on the rate of removal of
the MN nitrides by measurements of their number density.
As mentioned, calculations of the driving force for Z-phase precipitation allows estimations
of the precipitation rate of the Z-phase. The obvious strategy is then to minimise the driving
force in order to delay the harmful Z-phase precipitation to very long times (beyond 300,000
hours), where it would have no influence on the long-term stability of the steel in practical use.
Unfortunately, as shown this leads to limitations on the Cr content, which prevent attaining
the necessary oxidation resistance of the alloy.
However, if the Z-phase itself could be provoked to precipitate very quickly in a fine
distribution, which could provide particle strengthening, then the steel could be expected to be
immune to dissolution of its strengthening nitrides. Inspections of the developed
thermodynamic model show that the necessary steps to accelerate Z-phase precipitation
involve an increase of the Cr-content to the highest possible level. Further, the Carbon content
should be low in order to accelerate the Z-phase formation together with additions of Cobalt,
which would also limit delta ferrite. The high Cr content would then be turned from being the
main cause of instability into being a necessary beneficial element to secure the long-term
microstructure stability of the steels. And at the same time the necessary oxidation resistance
would be obtained, which would again make it possible to combine strength and oxidation
resistance in the same alloy.
Figure 5. Driving force for Z-phase precipitation in 9-12%Cr steels. Z650 is an experimental
12% Cr steel.
Scrutinising of the thermodynamic model suggests that Z-phases based on CrNbN or CrTaN
could possibly serve as strengthening agents, as the driving force for their formation are
significantly higher compared to that of CrVN or Cr(V,Nb)N, Fig. 5. Finally, calculations of
coarsening rates based on thermodynamic data made as described in [14], indicate coarsening
stabilities of the Z-phases similar to or better than the MN nitrides.
Fine NbN or TaN nitrides, which are precipitated during tempering, transform to a fine
distribution of Z-phase in relatively short time at 650°C. The strong acceleration of the Z-
phase transformation means that each individual MN nitride transforms quickly into one Z-
phase. Thereby the detrimental Z-phase growth process involving MN dissolution is
effectively suppressed. The result is a fine distribution of Z-phase particles, see Fig. 6, which
should be very stable against coarsening.
Figure 6. The principle in accelerated Z-phase transformation is to avoid growth of Z-phase,
and thereby obtaining a stable fine distribution of nitrides which contributes to creep strength.
6. Summary
The long-term creep strength of the new generation of martensitic creep resistant 9-12%Cr
steels since the well-known steel Grade 91 relies strongly on particle strengthening by fine
MN nitrides based on V and Nb. The MN nitrides may be replaced by the thermodynamically
more stable Z-phases, Cr(V,Nb)N causing a breakdown in long-term creep strength.
Thermodynamic modelling show that Cr contents above 10.5% strongly accelerates Z-phase
precipitation and the breakdown in creep strength. This explains the lack of success for all
attempts to develop martensitic creep resistant steels with high Cr content for oxidation
protection.
Careful control of the Z-phase precipitation process has led to the design of experimental
12%Cr martensitic steels strengthened by Z-phase. Such steels may again enable the
combination of high strength and oxidation resistance in the same alloy. This opens a new
pathway for further alloy development of the heat resistant martensitic steels.
If successful, the 12%Cr martensitic steels strengthened by Z-phase could enable the
construction of low cost USC power plants with steam parameters approaching 325bar/650°C
entirely based on steels.
Acknowledgments
Financial support from the power utilities DONG Energy and Vattenfall GN is kindly
acknowledged. The described work is part of the European COST 536 ACCEPT action and
the Swedish CROX project supported by the SSF (Stiftelsen för Strategisk Forskning) and the
VGB Research Foundation (Project no. 266).
References
[1] F. Masuyama: Proc. Conf. “Materials for Advanced Power Engineering 2006”, (ed. J.
Lecompte-Beckers et. al.), Part I, 175-188; 2006, Forschungszentrum Jülich.
[2] T.-U. Kern, M. Staubli, K. H. Mayer, B. Donth, G. Zeiler and A. DiGianfrancesco: Proc.
Conf. “Materials for Advanced Power Engineering 2006”, (ed. J. Lecompte-Beckers et. al.),
Part II, 843-854; 2006, Forschungszentrum Jülich.
[3] C. Henderson: “Case studies on recently constructed fossil-fired plants and summary of
status of gasification combined cycle technology”; 2007, Paris, IEA/OECD.
[4] G. Scheffknecht, Q. Chen and G. Weissinger: Proc. Conf: “Parsons 2003, Engineering
Issues in Turbine Machinery, Power Plant and Renewables“ (ed. A. Strang et. al.), 112-128,
IOM3, 2003, London.
[5] J. Hald: VGB PowerTech 12, 74 (2004).
[6] T. Fujita: Metall. Trans. A 12, 1071 (1981).
[7] H. K. Danielsen and J. Hald: Energy Mater. 1, 49 (2006).
[8] A. Strang and V. Vodarek: Materi. Sci. Tech 12, 552, (1996).
[9] F. Abe: Current Opinion in Solid State and Materials Science 8, 305 (2004).
[10] S. Straub: Verformungsverhalten und Mikrostruktur warmfester martensitischer 12%-
Chromstähle, 405, VDI Forschungsberichte 5, VDI Verlag, 1995, Düsseldorf.
[11] P. Polcik: Modellierung des Verformungsverhalten der warmfesten 9-12% Chromstähle
in Temperaturbereich von 550-650°C, 92, Shaker Verlag, 1999 Aachen.
[12] J. Hald: Proc. 1st Int. Conf. “Super-high strength steels”, AIM and CSM, Paper 3.4. 2005,
Rome.
[13] F. Abe: Sci. Technol. Adv. Mater. 9, 1 (2008).
[14] J. Hald and L. Korcakova: ISIJ Int. 43-3, 420 (2003).
[15] F. Abe: Int. J. Mat. Res. 99, 387 (2008).
[16] M. Hättestrand and H.-O. Andrén: Mater. Sci. Eng. A, 270, 33 (1999).
[17] E. Schnabel, P. Schwaab and H. Weber: Stahl Eisen 107, 691 (1987).
[18] H. K. Danielsen and J. Hald: Comput. Coupling Phase Diagrams Thermocem. 31, 505
(2007).
[19] F. Abe: Proc. Int. Conf. on New Developments on Metallurgy and Applications of High
Strength Steels, 2008, Buenos Aires.
[20] NIMS Creep Data Sheet, No 51, National Institute For Materials Science, 2006, Japan.
[21] H. Naoi, M. Oghami, Y. Hasegawa and T. Fujita: Proc. Conf: “Materials for Advanced
Power Engineering 1994 (ed. D. Coutsoradis et. al.), Part I, 425-434, Kluwer, 1994,
Dordrecht.
[22] T. Fujita, Personal Communication, 2008.
[23] K. Sawada, H. Kushima, K. Kimura and M. Tabuchi: ISIJ Int. 47, 733 (2007).
[24] H. K. Danielsen: “Z-phase in 9-12%Cr steels” Ph.d. Thesis, 2007, Technical University
of Denmark.
[25] H. K. Danielsen and J. Hald, Mater. Sci. Eng. A 505, 169 (2009).
[26] R. Blum and R. W. Vanstone: Proc. Conf. “Materials for Advanced Power Engineering
2006”, (ed. J. Lecompte-Beckers et. al.), Part I, 41-60, 2006, Forschungszentrum Jülich.
[27] H. Tschaffon: Proc. Conf. “Materials for Advanced Power Engineering 2006”, (ed. J.
Lecompte-Beckers et. al.), Part I, 61-68; 2006, Forschungszentrum Jülich.